diff options
author | Iwan Kawrakow <iwan.kawrakow@gmail.com> | 2024-06-26 17:38:18 +0300 |
---|---|---|
committer | Iwan Kawrakow <iwan.kawrakow@gmail.com> | 2024-06-26 17:38:18 +0300 |
commit | 0a3a2c4cd47943dc4c3c43be75728402584a3732 (patch) | |
tree | 9e3b58646f002f023de8991dc8b30f2d0251796a /common/common.h | |
parent | 71725a918f9edee559a978397779486dce7c703a (diff) |
imatrix: be able to specify the name of the output tensor
For some models the same tensor is used for token embeddings and
output. This tensor tends to be named token_embedding.weight rather
than output.weight, which prevernts us from collecting imatrix data
for this tensor. With this commit we can tell the name of the
output tensor to the imatrix tool.
Diffstat (limited to 'common/common.h')
-rw-r--r-- | common/common.h | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/common/common.h b/common/common.h index 9a1dc4a2..bb45b3b4 100644 --- a/common/common.h +++ b/common/common.h @@ -224,6 +224,7 @@ struct gpt_params { // imatrix params std::string out_file = "imatrix.dat"; // save the resulting imatrix to this file + std::string output_tensor_name = "output.weight"; // name of the output tensor int32_t n_out_freq = 10; // output the imatrix every n_out_freq iterations int32_t n_save_freq = 0; // save the imatrix every n_save_freq iterations |